Bi-objective optimization using an improved NSGA-II for energy-efficient scheduling of a distributed assembly blocking flowshop

In this study, an Energy-Efficient Distributed Assembly Blocking FlowShoP (EEDABFSP) is considered. An improved Non-dominated Sorting Genetic Algorithm-II (NSGA-II) is developed to solve it. Two objectives have been considered, i.e. minimizing the maximum completion time and total energy consumption. To begin, each feasible solution is encoded as a one-dimensional vector with the factory assignment, operation scheduling and speed setting assigned. Next, two initialization schemes are presented to improve both quality and diversity, which are based on distributed assembly attributes and the slowest allowable speed criterion, respectively. Then, to accelerate the convergence process, a novel Pareto-based crossover operator is designed. Because the populations have different initialization strategies, four different mutation operators are designed. In addition, a distributed local search is integrated to improve exploitation abilities. Finally, the experimental results demonstrate that the proposed algorithm is more efficient and effective for solving the EEDABFSP.